By David Wojick
This article is about questions not answers. When it comes to understanding the electric power grid there is a big data gap that needs to be overcome. We have no actual data on how much electricity EVs consume or when and where they consume it.
The problem is that most EVs most of the time are charged at the owner’s residence or workplace. In both cases the electricity consumed is often simply added to that building’s total on the meter. There is no separate accounting for it so no data. There are about 7,000,000 EVs on the American road with unknown grid impacts.
We do have a government guess to a preposterous eight significant figures, based on a computer model. But there is no way to test or calibrate the model so there is no reason to believe its output.
Still we can use these conjectured numbers to make some basic points about EV power and energy usage. We will treat them as true for the moment. They are from EIA here.
Estimated total EV energy consumption for the last three years is this:
2025 = 23,532,855 MWh 2024 = 17,800,214 MWh 2023 = 13,212,000 MWh
Note that the number almost doubled in just two years. That is remarkable growth especially when people talk about the EV market being dead.
Still even the biggest number is just a tiny fraction of U.S. electric energy consumption. As a result many articles say it shows that EVs have little or no grid impact. That conclusion is seriously wrong because there is more to the grid than total annual statistics.
What mostly challenges the grid is power not energy, MW not MWh, especially peak demand. Here EVs can have a significant impact because they are in effect peak demand generators.
The typical EV only charges for a very small fraction of the time so the charge rate is quite large. In fact there is a tremendous effort to develop ever faster charging which makes the charge peak even bigger.
When lots of EVs charge at the same time the combined peak demand can be very significant. It is the impact of a lot of EVs charging at the same time that we need to understand.
Here is a wildly worst case scenario that that makes the potential impact clear. First let’s assume all the 7 million or so EVs charge at the same time. Second let’s say they charge three times a week and it takes just a half hour each time, with a constant charge rate. (This short time is reasonable because most will just be topping off to get a full charge.)
Taking the 2025 number at a round 24,000,000 MWh, that amount is drawn from the grid in 156 hours. The charge rate is just under 154,000 MW.
This is a huge demand. In fact it is roughly equal to the record peak demand at PJM which serves 67 million customers. Demand like this would quickly fry the grid.
Of course the reality is far more complex and nothing even remotely like this has ever happened. But there are only so many hours a day and most of those 7 million EVs have to be charged fairly often.
The big question is what do the potential many-EV combined peaks look like at a realistic scale? They may already be a problem at the local level.
While the reliability problem may yet be small the cost problem might already be big. It has long been claimed that many local power distribution networks will need to be upgraded to provide the capacity to handle EV charging.
Upgrading is expensive and if there is a lot going it it might be a significant component of the present price increases. Is there a correlation between EV registrations and electricity price increases?
The fact that EV power usage is invisible could be hiding problems that are serious and rapidly growing. We really need to understand what is going on.
